|
It is therefore interesting to note that some of the most successful value investors of the past and present, on whom Leithner & Company’s approach to investment is based, pay no attention whatsoever to experts’ predictions and forecasts. Benjamin Graham, for example, stated that “the farther one gets away from Wall Street, the more scepticism one will find about the pretensions of stock-market forecasting or timing.” Peter Lynch, one the most successful funds managers of the last four decades, has disclaimed any ability to predict financial markets’ level or direction. Lynch has stated that market strategists “can’t predict markets with any useful consistency, any more than gizzard squeezers could tell the Roman emperors when the Huns would attack.” Similarly, Warren Buffett places no confidence in market forecasts and wastes no time on economic predictions: “if Fed Chairman Alan Greenspan were to whisper to me what his monetary policy was going to be over the next two years, it wouldn’t change one thing I do.”
How Accurate Are Forecasts Anyway?
These eminent investors have good reason to dismiss economic and market forecasts as well as the experts who make them. This is because economists’ ability to predict inflation, interest rates and the like is at best very limited; and the ability of analysts and “strategists” to predict either market indices or companies’ earnings, share prices and so on is even worse. This statement should not be taken as a criticism directed solely at economists: William Sherden, in his outstanding book The Fortune Sellers(a must-read for all serious investors), reviews research about the accuracy of various experts’ (including demographers and management consultants) forecasts, and finds that their accuracy is universally poor.
Sherden concludes that economists cannot, as a rule, predict turning points in economic series. Between 1980 and 1995, the U.S. Federal Reserve Board predicted three of the six turning points in GNP growth and neither of the two turning points in inflation. More recently, virtually all elite economic forecasters – including the IMF, OECD and international ratings agencies – failed to foresee the Asian economic crisis of 1997; and once it occurred, few if any domestic forecasters accurately predicted its impact upon the Australian economy.
Sherden also finds that economists’ forecasting skill, on average, is no better than the “naïve forecast” that the near future will continue to be pretty much like the recent past. Moreover, increased sophistication (i.e., more powerful computers, more complicated econometric models and greater amounts of data) has not improved the accuracy of economists’ forecasts; there is no evidence that forecasters’ skill has increased over the last twenty or thirty years – if anything, it has deteriorated over time; “consensus forecasts” (forecasters, like everybody else, find safety in numbers) are no more accurate than individual forecasts; and there are no individual forecasters who are consistently more accurate than their peers.
With respect to analysts’ predictions of companies’ earnings, David Dreman, Forbes columnist and Chairman of Dreman Value Management, reviewed almost 100,000 earnings forecasts made between 1973 and 1996 and found that the average forecast errs by no less than 30-60%. The odds are 1 in 130 that an analyst can predict a company’s earnings to within 5% of its true value four times in succession. Not surprisingly, Dreman concluded that company earnings are “utterly unpredictable” and (seemingly as an understatement) that analysts’ predictions of earnings are of “not much value.” Sherden’s summary conclusion with respect to economic and financial market experts: despite their advanced methods, (often) incomprehensible jargon and (almost invariably) high salaries, they are no better at predicting the future than astrologers and fortune-tellers.
Why Do the Experts Predict So Poorly?
Three reasons. The first taints private sector forecasters; the second tars all forecasters; and the third gives public sector forecasters a disproportionately hard kick in the pants.
The first reason is that the purpose of many forecasts and forecasters is not to make accurate predictions: it is to sell things, e.g., to persuade investors to make particular investments with particular investment firms. According to Reserve Bank Assistant Governor Glen Stevens (quoted in September 1999 by Australian Financial Review correspondent Stephen Koukoulas), many people forecast “with a view to selling a product, or a piece of advice. many forecasts made in the private sector are essentially of this variety. The forecaster has a story to tell in order to provide credibility to their employer’s efforts to win business.” Perhaps this is why market analysts’ and strategists’ predictions tend to err in the direction of over-optimism.
Similarly, according to Timothy Vick, author of the excellent book Wall Street on Sale, “Wall Street exists to sell you something. All the financial information it issues, whether brokerage recommendations, price targets, market forecasts, earnings estimates or performance figures, can be twisted to serve the purposes of whoever issues it.” For this reason, Vick states that investors should be sceptical of almost all data and forecasts which experts shower upon them.
Second, as economists such as Nobel Laureate Friedrich von Hayek argue, individuals’ choices about the options available to them are inherently subjective. Different individuals, having different perceptions, amounts and types of information of varying quality, respond differently to particular economic events; moreover, over time the same individual will perceive things differently and therefore make different choices with respect to available options. Accordingly, the context in which they make these choices – the market – is not a machine which can be studied in the manner that an engineer studies a machine. Rather, markets are “complex systems” which have counter-intuitive cause-and-effect results and unintended consequences, exhibit periods of order punctuated by unexpected moments of turmoil, adapt to changing circumstances and – most importantly for investors to realise – have no fixed cycles.
In the view of Hayek and other Austrian School economists, the operations of complex systems involve far too many variables to enable anything approaching accurate prediction. Some forecasters implicitly recognise this. Chris Caton, BT Australia’s chief economist, has stated: “it’s disappointing but forecasts are about the future and the future is inherently unknowable.” For the same reason, the present is just as difficult to interpret. Respected Bloomberg columnist Caroline Baum puts the point in these terms: “to try and explain why markets do what they do on any given day is a Herculean effort. There are so many moving parts that it is almost impossible to quantify the short- and long-run shifts in the supply of and demand for a particular currency, commodity or financial instrument.” In short, there is no single, collective reason which prompts investors and speculators to buy and sell on any given day: there is no collective mood or psyche.
For this second reason, I believe whether it pertains to the present or future that caution should replace experts’ hubris. The hugely expensive efforts devoted to economic modelling and market forecasting (together with interventionist economic policies and social engineering more generally) are inherently misguided, wasteful, potentially quite destructive and should therefore be abandoned forthwith. In the words of influential investor Philip Fisher: “I believe that the economics which deals with forecasting business trends may be considered to be about as far along as was the science of chemistry during the Middle Ages. The amount of mental effort the financial community puts into this constant attempt to guess the economic future “makes one wonder what might have been accomplished if only a fraction of such mental effort had been applied to something with a better chance of proving useful.”
The third reason why predictions yield such inaccurate results is that the data which underlie them are (despite the integrity and best efforts of their compilers) less valid and reliable than most people seem to believe. It is not a case of “lies, damned lies and statistics”: rather, most government statistics must necessarily be estimates, and agencies revise these estimates over time as new and more complete information comes to hand. In some instances, economic statistics pertaining to a particular point in time undergo repeated revisions which require several years to complete. A significant amount of forecasting error is thus likely to stem from errors in the data themselves. As Oskar Morganstern wrote a generation ago in his classic work On the Accuracy of Economic Observations: “textbooks on national income and macro-economics show little if any evidence of the awareness of their data’s difficulties and limitations. In Great Britain, as in the U.S. and elsewhere, national income statistics are still being interpreted as if their accuracy compared favourably with the measurement of the speed of light.”
Subjectivism and errors in the data have an important consequence. Not only is the effort devoted to predicting and forecasting a waste of time: erroneous forecasts based upon faulty premises and data, which are routinely used by government policy makers, business executives, investors and consumers, will clearly tend to produce incorrect – and potentially costly and injurious – decisions.
Indeed, fundamentally important decisions have been made at least partially on the basis of wildly inaccurate extrapolations from erroneous data. Consider as an example the U.S. Presidential election of 1992, at which Bill Clinton defeated George Bush. Two premises underlay Clinton’s key message to voters (“It’s the economy, stupid”): economic growth was sluggish and President Bush was responsible for its sluggishness. Yet at precisely the moment that Clinton was successfully assailing the Bush Administration’s allegedly poor economic performance, the U.S. economy was growing at a blistering rate of 5.7 per cent. Unfortunately for Mr Bush, however, this fact was unknown during the election, and not until 1994 did U.S. authorities complete their revisions to key economic indicators for the period 1990-93. When they did so, they increased their estimate of real GDP by 25% and of real disposable income by a whopping 70%.
These problems which inhere in aggregate economic data are hardly confined to the U.S. Most Australian retailers, including its largest retailers, stated in late 1998 that the Christmas period would be one of their best ever. Economists thus predicted that official retail trade statistics for December would reflect retailers’ enthusiasm. It came as a great surprise when the December data showed one of the largest monthly decreases for some years in seasonally-adjusted retail trade. The decrease was so great that it depressed the month-by-month trend in retail sales so much that the new trend suggested that a sharp deceleration in retail trade – and possibly a recession – was in the cards, and thus prompted demands to cut official interest rates. Indeed, the December decrease was so great that the Australian Bureau of Statistics, the compiler of these figures, seemed to cast doubt upon their veracity. Confidence in their accuracy was not completely restored when the ABS announced a record (5.2%) seasonally-adjusted increase in retail trade for January 1999.
Decisions which rely upon inaccurate forecasts will tend not only to be erroneous and costly; rather than decrease uncertainty (which is the key rationale for forecasts) they can actually amplify it. The automobile industry, for example, like many manufacturing operations, has long production cycles and lead times. If auto executives relied solely on economic forecasts to set their production plans, they would incur huge costs, adding new capacity and increasing inventories in response to rosy projections, and shutting plants and cancelling contracts with suppliers in response to gloomy projections. Instead of relying on forecasts – and as management guru Peter Drucker suggested as early as the 1950s – managers in this industry treat forecasts cautiously and embrace common sense by putting in place robust plans which remain viable under very different (and quickly-changing) conditions. Investors would be very well advised to plan their investments in precisely the same manner.
What Can One Do?
As Director of Leithner & Co., how do I respond to the issues and difficulties raised in this report?
First, I keep firmly in mind that, despite its undoubted importance, economics is not an exact science which is capable of making accurate forecasts. For every economist, “market strategist” and the like there will be an equal and opposed economist – both of whose predictions are likely to be wrong. I therefore discount forecasts and forecasters, disclaim for myself any ability to predict the future and insofar as possible disengage myself from the emotions unleashed by economic and market uncertainty. According to Bridgewater Associates, a reputable American firm, investors waste much time and make significant mistakes as a direct consequence of collecting, interpreting, worrying about and acting on forecasts. “Squinting at biases and comments by [Federal Reserve] Governors is not a good way to get a sense for where interest rates are going. A lot of market noise is created by the circular dance of market participants trying to guess how the Fed is going to react to news. we prefer not to tango.”
Second, I recognise that to disclaim any ability to divine the future with any useful degree of accuracy is not to ignore the future. Quite the contrary: I plan for the future by considering scenarios – and particularly pessimistic scenarios – of what might happen (as opposed to predictions of what will happen), and structure Leithner & Co.’s investment portfolio accordingly.
Finally, I focus my attention upon neither the macroeconomy nor financial markets as a whole; rather, I concentrate upon individual companies and their securities. A small number of companies have stable characteristics whose future performance (in the sense of a naïve forecast) can be very tentatively projected on a “pessimistic scenario” basis. And most of all, I keep uppermost in mind that since the 1930s value investors such as Benjamin Graham, Walter Schloss and Warren Buffett have invested extraordinarily successfully and created significant wealth for their shareholders by basing their investment decisions upon thorough analysis of companies’ financial statements and keeping their own counsel. No role models, in my view, could be more useful.
 
Designed & maintained by
Artist Web Design
©1999-2008 All Rights Reserved |
